We examine an important case study of the controversial use of the initiative process in California: the colorful “Ham and Eggs” movement of 1938, a misguided attempt to provide elderly Californians with a weekly pension. Our analysis focuses on a collection of postcard survey responses recently discovered at the Huntington Library in Pasadena, California. This survey, conducted in 1938 by the Research Institute, posed questions about a series of pressing issues, including the various Ham and Eggs initiatives. We provide a qualitative study of the open-ended responses provided by survey participants along with a quantitative analysis of the fixed-choice questions. As best as we are aware, this is the only survey data on California politics of this period. Thus our analysis, while necessarily limited, sheds additional light on this important period of California politics.

In this paper I analyze a 3-person voting game in which players sequentially choose amendments to a bill and then vote sophisticatedly on the resulting agenda. This game differs from previous work on endogenous agenda formation in that the number of items on the agenda is not fixed. Players keep making proposals until they can do no better. I find that there is a nonempty set of minimax equilibrium outcomes to the game, and that this set contains the simple von Neumann-Morgenstern stable set.

In this paper, we model a legislative calendar as an ordered list of issues to be considered in sequence. We present two equilibrium models of legislative scheduling, one presuming that the calendar is itself the object of collective choice and the other considering which calendars are immune from “discharge” (the change of at most one element of the calendar by a decisive coalition in the legislature). We then examine the question of designing a scheduling process consistent with a preexisting decisive structure. We show that, so long as the status quo policy is Pareto efficient, then for any legislative calendar there exists a nontrivial scheduling process under which the calendar in question is stable. The results have implications for the robustness of structure induced equilibria in general and the design of intralegislative “gatekeepers” (such as the Rules Committee in the U.S. House of Representatives) in particular.

This paper considers manipulation of collective choice – in such environments, a potential alternative is powerful only to the degree that its introduction can affect the collective decision. Using the Banks set (Banks [1985]), we present and characterize alternatives that can, and those that can not, affect sophisticated collective decision making. Along with offering two substantive findings about political manipulation and a link between our results and Riker’s concept of heresthetic, we define a new tournament solution concept that refines the Banks set, which we refer to as the heresthetically stable set.

I present a formal model of the effect that representation can have on the formation of group identities using the debates over the drafting of the United States Constitution as a case study. I first show the presence of “factions,” or groups with competing interests, to be beneficial in forging a national identity. Next, I use this model to argue that the Great Compromise succeeded asmore than a political maneuver to ensure ratification of the Constitution; it created a political environment in which an American national identity could emerge. More generally, I find that representation schemes that ignore group distinctions and use the individual as the basic unit of political representation may induce individuals to embrace a group-based notion of identity. Conversely, acknowledging group distinctions by using the group as a unit of political representation may induce individuals to embrace a more universalistic conception of identity, and thus may make group distinctions less salient.

This paper examines the role that political institutions play in inducing or subduing ethnic identification. I provide a formal model of ethnic identification as a function of institutional design and social context that focuses on one question: What kinds of institutions will induce individuals to identify with the state rather than along ethnic lines? Among other things, the model offers predictions about when political institutions are capable of inducing a national identity. I conclude with an application of the model to the question of institutional design in Iraq. The model provides one explanation for why a period of low Sunni representation in Iraqi government would increase sectarianism, and why a more proportional Iraqi parliament may not remedy the problem.

In this paper, we define a binary relation for general collective choice situations, the needing relation, that captures one notion of dependence between alternatives with respect to inclusion in several popular solution concepts, including the top-cycle set, the uncovered set, the Banks set, and the minimal covering set. After defining the relation and relating it to a graph theoretical view of collective preference, we prove several results about the relation and provide a partial characterization of external stability of some tournament solution concepts. We then utilize the needing relation to define a property of tournament solution concepts, which we dub independence. Finally, the independence of several solution concepts is characterized.

I present a new method of interpreting voter preferences in settings where policy remains in effect until replaced by new legislation. In such settings voters consider not only the utility they receive from a given policy today, but also the utility they will receive from policies likely to replace that policy in the future. The model can be used to both characterize long-term preferences and distributions over policy outcomes in situations where policy is ongoing and voters are farsighted.

I provide a definition of the Banks set, or set of sophisticated voting outcomes, over an infinite policy space and when individual preferences are weak. I also show that the Banks set is a subset of one definition of the uncovered set, but not another. The interpretation of the Banks set in this setting differs from Banks’s original interpretation in the implicit role of the agenda setter. In addition, a characterization of the Banks set is provided for a three-player game of distributive politics. In this special setting, the Banks set and all definitions of the uncovered set have full measure over the space of alternatives.

Different definitions of the uncovered set are commonly, and often interchangeably, used in the literature. If we assume individual preferences are strict over all alternatives, these definitions are equivalent. However, if one or more voters is indifferent between alternatives these definitions may not yield the same uncovered set. This note examines how these definitions differ in a distributive setting, here each voter can be indifferent between any number of alternatives. I show that, defined one way, the uncovered set is equal to the set of Pareto allocations that give over half the voters a strictly positive payoff, while alternate definitions yield an uncovered set that is equal to the entire Pareto set. These results highlight a small error in Epstein (1998) in which the author characterizes the uncovered set for a different definition of covering than claimed.

I construct a theoretical and computational model of municipal fragmentation and Tiebout competition and show that factors such as income heterogeneity, political institutions and the discretionary power of municipalities to tax can substantially affect both where individuals choose to live and how cities form. Conclusions are drawn about the types of cities that form when secession is an option. These conclusions support the idea that increasing the range of choices available to municipalities and to individuals can actually leave a majority of residents worse-off.